COMPETING MARKETABILITY DISCOUNT METHODOLOGIES

Transcription

1 COMPETING MARKETABILITY DISCOUNT METHODOLOGIES A Review of the FMV Restricted Stock Study Z. Christopher Mercer, ASA, CFA, ABAR EXCERPTED FROM Valuing Shareholder Cash Flows: Quantifying Marketability Discounts (2005 E-Book)

2 Valuing Shareholder Cash Flows Quantifying Marketability Discounts 2005 E-Book 2005 E-Book Edition by Z. Christopher Mercer, ASA, CFA Published by Peabody Publishing, LP

3 Chapter 5 Competing Marketability Discount Methodologies THE FMV OPINIONS RESTRICTED STOCK STUDY TM The FMV Opinions Restricted Stock Study TM ( FMV Study ) was introduced in 2001 in an article written by employees of FMV Opinions, Inc. appearing in Valuation Strategies. 100 The introduction provided little substantive information about the contents of the study. To the best of our knowledge, other than articles written by employees of Mercer Capital, there has been no other review of the FMV Study by appraisers. Since then, there have been a number of articles by employees of FMV Opinions, Inc. yet these articles have also provided a very low level of insight into the nature of the FMV Study. Historical Development of Marketability Discount Analysis There have been several stages in the historical development of marketability discount analysis. These stages were outlined in a recent article in the Business Valuation Review (and summarized below): The 35% Stage. In the 1970s and 1980s, appraisers tended to consider marketability discounts in the range of 35%, based primarily on the averages of restricted stock studies summarized in the first two editions of Pratt s Valuing a Business. 2. The 35% to 45%, Plus or Minus, Stage. By the late 1980s, the combination of restricted stock studies and the pre-ipo studies (offered by John Emory, then at Robert W. Baird & Co., and by Willamette Management Associates), yielded average indications in the general range of 35% to 45%. Appraisers tended to believe that marketability discounts should be in that general range, and the venturesome occasionally extended the range, plus or minus, a little bit. 100 Robak, Espen and Hall, Lance S., Bring Sanity to Marketability Discounts: A New Data Source, Valuation Strategies, July/August Mercer, Z. Christopher, and Harms, Travis W., Marketability Discount Analysis at a Fork in the Road, Business Valuation Review, December 2001 (Volume 20, No. 4), pp Quantifying Marketability Discounts E-Book Page 154

4 3. Mandelbaum Benchmark Analysis Stage. Benchmark analysis came into its own in 1995 when Judge Laro correlated much of what appraisers had been doing into a nine factor analysis in Mandelbaum. 102 Some appraisers added factors and many used a form of the analysis advanced in Mandelbaum. However, like with the previous stages, the problem remained in that it was difficult to differentiate the facts and circumstances of a private investment relative to the information contained in the various averages. 4. The More Data Stage. By the early 1990s, a few appraisers acknowledged the issues with standard benchmark analysis, and began to conduct additional restricted stock studies with more detailed comparative information regarding transactions. The first of these studies, a study prepared by Management Planning, Inc. ( the MPI Study ), was published in Quantifying Marketability Discounts in Detailed transactional information for 49 transactions dating from 1980 to 1995 was provided in the MPI Study. In 2001, FMV Opinions, Inc. announced their study. As mentioned previously, few details of the study were provided in the Valuation Strategies article, but the study was offered for sale in an advertisement in the same issue. In that article, the authors suggested that the QMDM was a black box method and that it was flawed by circular reasoning. It would add little to this article to rehash the ensuing debate, but readers are referred to the resulting follow-up in Valuation Strategies. 103 Suffice it to say, as shown above, the QMDM is no more a black box than is any other discounted cash flow model and the reasoning behind the QMDM is no more circular than that of any discounted cash flow method. At the present time, the FMV Study, which is available for licensing through Pratt s Business Valuation Resources, contains two pieces. The first piece is the two year holding period study, containing data on 248 transactions which occurred between 1980 and 1997, when the Securities and Exchange Commission changed Rule 144 to lower the holding period in relevant restricted stock transactions from two years to one year. The second piece contains data on 182 restricted stock transactions occurring between 1997 and The FMV Study will be reviewed in considerable detail below. 5. The Quantitative (Rate of Return) Analysis Stage is Emerging. The use of quantitative analysis to develop marketability discounts has been growing since the mid-1990s (from only isolated usage prior) and accelerated following the publication of Quantifying Marketability Discounts in With this brief introduction to historical marketability discount analysis, we proceed to review the FMV Study. 102 Mandelbaum v. Commissioner, T.C. Memo , 69 T.C.M. (CCH) 2852 (1995), aff d., 91 F 3d (24(3d Cir. 1996)). 103 See Mercer, Z. Christopher and Heinz, Nicholas J., Marketability Discounts: Back to Reality, Valuation Strategies, November/December 2001, p. 34, in which we refuted these suggestions in detail. See also the Counterpoint debate in the July August 2002 issue of Valuation Strategies: Hall, Ekstein and Robak, The Case of the Black-Box Model vs. Empirical Data, and Mercer, QMDM: QED. Quantifying Marketability Discounts E-Book Page 155

5 Review of the FMV Study As noted above, the FMV Study was introduced for sale by FMV Opinions, Inc. in Since that time, the study has been licensed exclusively through Shannon Pratt s Business Valuation Resources. 104 To our knowledge, the study has not been independently reviewed and analyzed until now. 105 In an article, Lance S. Hall, ASA of FMV Opinions, Inc. discussed the need for better analysis in order to support and to defend marketability discounts in Tax Court. 106 The article reviewed four recent cases in the Tax Court in which a new approach developed by Dr. Mukesh Bajaj to determine discounts for lack of marketability has been given weight by the Tax Court. After pointing out several flaws with what he termed the Bajaj approach, Mr. Hall suggested that more data is the answer. The Bajaj approach uses only 40 private placements of restricted stock. Obviously the more data that is available, the better the analyst is able to perform a comparative analysis. Moreover, the Bajaj approach examines only four company characteristics. One study, the FMV Opinions Restricted Stock Study TM, has more than 240 private placement transactions for the period 1980 to 1996 and more than 190 private placement transactions for the period 1997 to In addition, the FMV Study looks at more than 30 company or transaction characteristics. Moreover, the Tax Court in both McCormick and Lappo indicated a strong preference for studies that have been performed by the testifying expert. With such data as supplied in the FMV Study, the appraiser can choose which characteristics to examine and which companies to include in the discount analysis; the FMV Study just supplies him or her with the necessary back-up data. Moreover, a significantly larger restricted stock data base with much greater company and transactional characteristics than the Bajaj approach s database is available to appraisers to make a comparative analysis and assessment of the appropriate discount for lack of marketability applicable to a specific privately held company. 104 As of this writing, the price for a one year license is $499 and for a two year license $918 (with no indication of any marketability discounts) In order to review the FMV Study, we purchased it. 106 Hall, Lance S., Counteracting the New and Winning IRS Approach to Determine Discounts for Lack of Marketability, Valuation Strategies, March/April 2004, p. 14. Quantifying Marketability Discounts E-Book Page 156

6 Guideline Company Analysis It is clear that Mr. Hall is suggesting that the FMV Study can be used by appraisers when employing the guideline company method to value illiquid minority interests in private companies. The Business Valuation Standards of the American Society of Appraisers provide Statements on Business Valuation Standards (SBVS). The SBVS develop specific guidance for certain topics. The first such statement addresses the Guideline Company Valuation Method. 107 SBVS-1, Paragraph II.C. states: Guideline companies are companies that provide a reasonable basis for comparison to the investment characteristics of the company being valued. Ideal guideline companies are in the same industry as the subject company; however, if there is insufficient transaction evidence available in that industry, it may be necessary to select other companies having an underlying similarity to the subject company in terms of such relevant investment characteristics as markets, products, growth, cyclical variability, and other salient factors. SBVS-1, Paragraph V. states: A comparative analysis of the qualitative and quantitative similarities and differences between the guideline companies and the subject company must be made to assess the investment attributes of the guideline companies relative to the subject company. Interestingly, SBVS-1 does not address the issue of the proximity in time between guideline transactions used for valuation inferences and the valuation date for the appraisal. However, the Definitions of Business Valuation Terms in the ASA s Business Valuation Standards define the valuation date as: 108 Valuation Date the specific point in time as of which the valuator s opinion of value applies (also referred to as Effective Date or Appraisal Date ). 107 American Society of Appraisers, Business Valuation Standards, SBVS-1 The Guideline Company Valuation Method, p. 31 (downloaded from the American Society of Appraisers website May 2004). 108 Ibid, p. 21. Quantifying Marketability Discounts E-Book Page 157

7 Groups of guideline public companies are routinely developed by appraisers when valuing private companies. Revenue Ruling 59-60, in fact, mandates the consideration of similar public companies when valuing private companies. A common characteristic of guideline public company analysis is that the pricing of public transactions is considered as of the valuation date, or certainly as close to the valuation date as is practicable given data limitations. However, for most public companies, daily trading information is available, and guideline groups are typically priced as of the valuation date. 109 The FMV Study and Time The FMV Study, as noted above, is divided into a two year study period, during which the SEC Rule 144 period of restriction was two years (until April 29, 1997), and a one year study period containing transactions occurring after the change in Rule 144 to a one year period of restriction. Transactions in the two studies occur as follows in Figure 5-1: 110 The FMV Study Two Year Portion One Year Portion Year Total Year Total Total Total 248 Figure Some appraisers do, however, examine the pricing patterns in the periods leading up to the valuation date for perspective and to spot unusual trading patterns that might disqualify a particular guideline company from consideration. 110 The entire analysis of the FMV Study is based on the data base as it existed in May Quantifying Marketability Discounts E-Book Page 158

8 The majority of the transactions in the two year portion of the FMV Study occurred in the 1990s. The majority of the transactions in the one year portion of the study occurred before the meltdown in the market following the tech boom period. No transactions have been reported on since It would seem that the use of transactions from the FMV Study as direct guideline company observations would not meet the customary standard of proximity to the valuation date for appraisals as of a current date. The FMV Study and Industry The two portions of the FMV Study can also be stratified by industry classification. One data point in the study is the four-digit SIC Code for each transaction. For purposes of this review, we have sorted the transactions by two digit codes for a broader look (see Figure 5-2). Nine industry classifications account for 185 (75%) of the 248 total transactions in the two year portion of the study. The remaining 63 transactions are spread over 25 other industry categories. Interestingly, of the 28 transactions in SIC Classification #28, 18 represented companies in SIC Code #2834 (pharmaceutical preparations). Of the 24 transactions in SIC Classification #87, 15 are in SIC Code #8731 (services, commercial, physical and biological research). The FMV Study Two Year Portion Two Digit SIC Codes No. Two Digit SIC Classification Description Measuring, Analyzing & Controlling Instruments Chemicals & Allied Products Electronic & Other Electrical Equip & Components, except computers Engineering, Accounting, Research, Management & Related Svcs Business Services Holding & Other Investment Ofices Industrial & Commercial Machinery & Computer Equipment Oil & Gas Extraction Communications 25 Others 63 Various Total 248 Figure 5-2 The one year portion of the FMV Study can be similarly sorted (see Figure 5-3). Five industry classifications account for 139 (76%) of the 182 transactions, with the 43 remaining transactions being spread over 17 other industry classifications. Of the 54 transactions in SIC Classification # 73, 29 are in SIC Code #7372 (services, prepackaged software), and 20 are in SIC Code #7375 (services, computer-related). Further, of the 26 transactions in SIC Classification #28, 14 are in SIC Code #2834 (pharmaceutical preparations). Of the 22 transactions in SIC Classification #87, 20 are in SIC Code #8731 (services, commercial, physical and biological research). Quantifying Marketability Discounts E-Book Page 159

9 The FMV Study One Year Portion Two Digit SIC Codes No. Two Digit SIC Classification Description Business Services Measuring, Analyzing & Controlling Instruments Chemicals & Allied Products Engineering, Accounting, Research, Management & Related Svcs Electronic & Other Electrical Equip & Components, except computers 17 Others 43 Various Total 182 Figure 5-3 It should be clear from this brief classification analysis that the transactions in both the two year and one year portions of the FMV Study are concentrated in a relatively small group of industries. And this small group of industries is not generally representative of privately owned corporate America. 111 Given the thousands of publicly traded companies in the public securities markets where prices are determined every day in active or relatively active trading, it is quite often impossible to find a group of sufficiently comparable companies (to a given private valuation subject) with which to employ the guideline company method. With a total of only 430 transactions occurring across a span of 22 years, and concentrated in a relatively few industry classifications, the FMV Study provides a limited source for guideline company analysis. Quintile Analysis of the Two Year Portion of the FMV Study With a broad overview of the FMV Study s transactional database in mind, it is now appropriate to examine the transactions in more detail to understand the operating nature of the companies involved and any relationships that may be discernible with respect to operating characteristics and restricted stock discounts. 111 Companies whose transactions are in the FMV Study often engaged in multiple restricted stock transactions. In the two year portion, 34 companies engaged in a total of 86 of the 248 transactions. In the one year portion, 32 companies engaged in 76 of the total of 182 transactions. And the restricted stock discounts sometimes varied considerably when the same company engaged in multiple issues. For example, AT&E Corporation, engaged in the wristwatch message paging business, issued restricted stock in March and April of 1987, with one discount being 24% and the other being 50%. Quantifying Marketability Discounts E-Book Page 160

10 Limited analysis of the FMV Study was provided in a recent article. 112 The data base was divided into quintiles (such that there are five groups of equal size). We performed a similar analysis for this chapter, except that we created a sixth category including all transactions occurring at restricted stock premiums, or at prices above the freely traded prices of the respective companies. We made this adjustment because there are apparently different forces or factors at work with those transactions and we found it helpful to examine them separately. Before proceeding, we should note that the FMV data base includes pricing as of three dates. Prices for each company are recorded in the data base based on the average of the high and low prices for a) the month preceding the transaction, b) the month in which the transaction occurred, and c) the month subsequent to the transaction. The analyses used by FMV Opinions, Inc. in various publications appears to focus on discounts based on subsequent month pricing. However, Figure 5-4 indicates there is a difference in median discounts based on the three different pricing dates. Discounts for Two Year Portion PriorMonth TransMonth SubMonth Median 16.7% 20.0% 22.0% Average 18.2% 21.9% 22.5% Std Dev 16.7% 16.0% 21.5% Discounts for One Year Portion PriorMonth TransMonth SubMonth Median 11.7% 21.8% 23.0% Average 11.2% 24.1% 20.7% Std Dev 29.4% 22.4% 35.2% Differences in Discounts PriorMonth TransMonth SubMonth Median -5.0% 1.8% 1.0% Average -7.0% 2.2% -1.8% Figure 5-4 The median discounts based on subsequent month pricing are 22.0% for the two year portion of the study and 23.0% for the one year portion. This observation would suggest that discounts increased with a decrease in required holding periods, which is counter to theoretical expectations, assuming other things remain equal. This apparent anomaly was attributed to an increase in volatility between the two periods. 113 Note that the differences are as expected when prior month pricing (or even average discounts for the subsequent month). Perhaps, other things were not equal (other than volatility), or there were other reasons for higher volatility, as will be seen below. 112 Hall, Lance S., The Search for the Holy Grail: Getting Away from the 15-Minute Discount Determination, The Value Examiner, July/August Hall, Lance, S., Why Are Restricted Stock Discounts Actually Larger for One-Year Holding Periods? Business Valuation Update, September Quantifying Marketability Discounts E-Book Page 161

11 For the two year portion of the FMV Study, the median and average discounts rise based on the selection of pricing based on the high/low average of the prior month (which would presumably reflect no market reaction to the yet unannounced transaction), to the transaction month (which could be based on a price before or after the announcement), to the subsequent month pricing (which would incorporate the markets reaction to the deal). A similar trend is seen in the one year portion of the data base; however, the jump between prior month and transaction month discounts is much more pronounced in the one year portion. Users of the FMV Study should be aware of the sensitivity of median discounts to the selection of the pricing date for transactions. The following analysis was conducted using 14 of the more than 50 fields in the FMV Study s database. Figure 5-5 summarizes the quintile analysis and provides overall median observations for the two year portion of the study, as well. OVERVIEW ANALYSIS OF THE TWO YEAR ANALYSIS FMV OPINIONS RESTRICTED STOCK STUDY TM (248 Restricted Stock Transactions Occurring ) Transactions First Second Third Fourth Fifth Overall at Premiums Quintile Quintile Quintile Quintile Quintile Medians Median Restricted Stock Discounts -5.4% 4.9% 15.9% 25.0% 34.7% 51.0% 22.2% Number of Transactions Transaction Pricing Median Statistics for each Characteristic Per Share Price $11.53 $9.56 $6.00 $6.20 $4.00 $3.04 $6.00 Offer Amount ($mm) $9.6 $8.0 $5.9 $4.5 $2.6 $2.8 $4.7 % of Company Placed 10.3% 7.9% 12.5% 10.1% 8.7% 14.7% 10.8% Pre-Transaction Co. Stats ($mm) Market Capitalization $100.4 $115.4 $72.9 $46.8 $31.4 $25.4 $51.6 Book Value $8.0 $20.2 $5.60 $10.4 $4.5 $3.5 $7.0 Price/Book Multiple Total Assets $19.4 $38.7 $14.6 $23.5 $10.1 $6.9 $15.6 Revenues $15.0 $29.8 $11.9 $25.4 $9.8 $5.6 $12.8 EBITDA $0.8 $1.3 $0.3 $0.8 ($0.5) ($0.4) ($0.4) Net Income ($1.2) ($1.0) ($0.8) ($0.7) ($1.3) ($0.9) ($0.9) Operating Margin -0.4% 1.7% -4.5% 0.1% -8.0% -8.1% -2.8% Volatility Z-Score Dividend Yield 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Implied Market Cap/Revenues Figure 5-5 Quantifying Marketability Discounts E-Book Page 162

12 The overall median restricted stock discount for the pre-1997 transactions is 22.2%, as seen at the upper right portion of the table above. 114 As with all previous studies of restricted stock transactions, looking only at median discounts (the far right column in Figure 5-5) masks the broad range of discounts in the study. The actual discounts range from a high discount of 81% (for a transaction involving LXR Biotechnology, Inc. in December 1996) to a low discount, actually a 92% premium (for a transaction involving ACS Enterprises, Inc., a company in the analog and wireless television subscription businesses in October 1993). Some 23 transactions occurred at a premium discount with a median premium of 5.4% (a restricted stock discount of minus 5.4%), as seen in the far left column of Figure 6. Premium transactions can occur because the purchase price was actually set at a premium to the market price, or because of a decrease in the stock price between the date of the announcement and the subsequent month pricing in the FMV Study, or both. We do not really know the reasons for premium transactions, so they are treated separately. The remaining 225 transactions were divided into five quintiles of 45 transactions each. As indicated at the top-middle of the table above, there is a steady rise in restricted stock discounts across quintiles, with the first quintile having a 4.9% median discount and the fifth quintile having a median observation of 51.0%. We can observe the following from the quintile analysis: The median market capitalization declines steadily from $115 million (Q1) to $25.4 million (Q5). The obvious inference is that restricted stock discounts tend to increase as size, as measured by market capitalization, increases. 115 There is no similar correlation with size, as measured by revenues. The median company in every quintile is losing money. Only 98 companies of the 248 observations in the pre-1987 study were profitable, meaning that some 60% of the transactions involved companies that were losing money. Volatility (as measured by the annualized standard deviation of the continuously compounded rate of return on each company s common stock with the standard deviation determined by examining the week-over-week difference in the weekly closing price analyzed over the one year period prior to the transaction date) generally rises over the quintiles from 0.60 (Q1) to 0.85 (Q5). The median dividend yields are all 0%, and only 24 of the 248 transactions involved companies paying dividends. 114 The slight discrepancy in the quintile calculation and the overall median calculation presented above (22.2% versus 22.0%) results from the fact that in the quintile analysis, if there was no price in the database for the subsequent month, we attributed the transaction month discount for that observation (so that our quintiles would have even numbers of observations). 115 Despite this observed trend in the quintiles, there is no statistically significant correlation between observed discounts and either market capitalization or revenues (or of the log of each). Quantifying Marketability Discounts E-Book Page 163

13 The surprising characteristic of the median revenue analysis is that, overall, the companies in the two year portion are quite small by public standards (median revenues of $12.8 million). The implied multiples of market capitalization to revenues (based on median pricing) are surprising, with an overall median of 4.0x revenues. 116 The bottom line of this analysis is that the two year study is comprised of fairly small companies selling at quite rich market valuations. The median price/book multiple implied by market pricing was 5.8x, again indicating relatively rich market pricing in comparison to most closely held companies valued by appraisers. Quintile Analysis of the One Year Portion of the FMV Study We prepared a similar analysis of the one year portion of the FMV Study, as shown in Figure 5-6. OVERVIEW ANALYSIS OF THE ONE YEAR ANALYSIS FMV OPINIONS RESTRICTED STOCK STUDY TM (182 Restricted Stock Transactions Occurring ) Transactions First Second Third Fourth Fifth Overall at Premiums Quintile Quintile Quintile Quintile Quintile Medians Median Restricted Stock Discounts -20.0% 8.1% 20.3% 32.9% 45.8% 67.8% 23.0% Number of Transactions Transaction Pricing Median Statistics for each Characteristic Per Share Price $11.08 $8.43 $6.87 $7.50 $5.00 $3.38 $7.45 Offer Amount ($mm) $11.6 $11.3 $4.6 $5.0 $5.0 $3.3 $5.1 % of Company Placed 6.7% 8.7% 4.8% 6.4% 7.2% 12.1% 7.2% Pre-Transaction Co. Stats ($mm) Market Capitalization $181.0 $137.9 $95.2 $89.2 $88.7 $36.1 $104.3 Book Value $15.8 $30.1 $4.20 $8.40 $6.50 $5.6 $9.1 Price/Book Multiple Total Assets $28.2 $62.1 $7.8 $14.0 $10.4 $7.5 $18.5 Revenues $12.1 $20.5 $3.4 $4.4 $2.4 $1.8 $4.4 EBITDA na na na na na na na Net Income ($9.3) ($5.2) ($5.9) ($4.9) ($4.9) ($7.1) ($6.8) Operating Margin -69.2% -39.5% % -40.4% % % -97.3% Volatility Z-Score Dividend Yield 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Implied Market Cap/Revenues Figure 5-6 The overall median restricted stock discount for the one year portion of the FMV Study is 23.0% (based on subsequent month discounts), or slightly higher than the median for the two year portion of the study. As with the two year analysis, we can make a number of observations: 116 We calculated the preferred market value of total capital to revenue multiples for the two year study. The overall median multiple was 3.6x, or fairly close to the 4.0x multiple shown in the table. Quantifying Marketability Discounts E-Book Page 164

14 The discounts rise consistently across quintiles, with the fifth quintile median, at 67.8%, being considerably higher than the fifth quintile median of the two year portion (51.0%). This result occurred (with similar overall medians) at least in part because of the higher proportion of observations exhibiting negative discounts (restricted stock premiums). 43 (24%) of the 182 transactions reflected restricted stock premiums. Median discounts for the quintiles are again negatively (visually) correlated with size as measured by market capitalization. The median company measured by revenues has only $4.4 million in revenues. By way of perspective, 12 companies had no revenues, another 12 companies had revenues less than $100 thousand, and a total of 43 companies (inclusive) had revenues less than $1.0 million. The median observations for all quintiles indicate substantially negative profitability. Only 26 of the 182 observations involved companies with profitable operations (defined as pre-tax greater than zero). Volatility generally rose across the quintiles, with an overall median volatility of The median dividend yield was 0%. None of the companies in the one year portion of the FMV Study paid dividends. The surprising characteristic of the median revenue analysis for the more recent, one year study is that, overall, like the two year study, the companies in the one year portion are quite small by public standards (median revenues of only $4.4 million). The implied multiples of market capitalization to revenues (based on median pricing) are astonishing, with an overall median of 23.7x revenues. The bottom line of this analysis is that the one year study is comprised of quite small companies selling at very dear market valuations. The median price/book multiple implied by market pricing was 7.6x, again indicating rich market pricing in comparison to most closely held companies valued by appraisers. Comparing the Medians Two Years vs. One Year To compare the results of the two year portion with the one year portion, we provide a summary table in Figure 5-7 highlighting the differences in median observations. Recall that the Hall article attributed the increase in the one year median restricted stock discount to an increase in the volatility of the earnings of the companies in the two portions of the study. Quantifying Marketability Discounts E-Book Page 165

15 COMPARISON OF MEDIAN OBSERVATIONS FMV OPINIONS RESTRICTED STOCK STUDY TM TWO YEAR PORTION VS. ONE YEAR PORTION Two Year One Year Overall Overall Medians Medians Differences Observations Restricted Stock Discounts 22.2% 23.0% 0.8% "Significant" implies Number of Transactions change > 20% Transaction Pricing Per Share Price $6.00 $7.45 $1.45 Significant increase Offer Amount ($mm) $4.7 $5.1 $0.4 Increase % of Company Placed 10.8% 7.2% -3.6% Significant decrease Pre-Transaction Co. Stats ($mm) Market Capitalization $51.6 $104.3 $52.7 Significant increase Book Value $7.0 $9.1 $2.1 Significant increase Price/Book Multiple Significant increase Total Assets $15.6 $18.5 $2.9 Increase Revenues $12.8 $4.4 -$8.4 Significant decrease EBITDA ($0.4) na nm Net Income ($0.9) ($6.8) ($5.9) Significant decrease Operating Margin -2.8% -97.3% -94.5% Significant decrease Volatility Significant increase Z-Score Significant increase Dividend Yield 0.0% 0.0% 0.0% Implied Market Cap/Revenues Significant increase Figure 5-7 Our analysis suggests that there are numerous reasons for any differences in the two median restricted stock discounts. Other things did not remain the same. There were significant changes (defined as 20% or more, plus or minus), in several important operating statistics and market pricing. Fueled by robust markets prior to the tech meltdown, the median market capitalization more than doubled between the two portions of the study. The portion of the company being sold was considerably smaller in the latter analysis. Relative pricing as measured by price/earnings multiples or price/book multiples increased significantly in the latter study. Profit dynamics also deteriorated considerably as measured by margins or by earnings. In short, the one year study (after April 1997) contains transactions involving companies even less similar to private corporate America than the pre-1997 transactions. Quantifying Marketability Discounts E-Book Page 166

16 Comparing Conclusions In a 2004 article, regarding a quintile analysis of the two year data, Mr. Hall writes: In assessing the relevance of these data, it is important to understand that private placements of restricted stock do not reflect the normal cross section of all publicly traded companies. Rather, pure play restricted stock transactions are heavily weighted towards the small, micro-cap segment of the public marketplace. Many of the attributes of the companies in the FMV Study are similar with many of the small privately held companies that appraisers typically value. 117 It should be clear that this analysis is incomplete and does not focus on the salient operating characteristics of the companies in the FMV Study. For example, the quintile analysis tables in the Hall article include only median data for restricted stock discounts, market value (capitalization), volatility, total assets, revenues, and price per share. The more complete analysis in this chapter does not suggest general comparability with private corporate America. Block Size vs. the Discount Mr. Hall states the following in reference to a graph in his article entitled Block Size vs. The Discount: A Closer Look At The Top Quintile : 118 The table above demonstrates that the marketplace recognizes the differences in liquidity between large percentage blocks of restricted stock and small percentage blocks of restricted stock. Percentage blocks less than 20% had a median discount of only 18%, whereas the median discount for blocks over 35% had a median discount of 58 percent. By recognizing the differences in liquidity between large- and small-blocks, the FMV Study data become meaningful in the valuation of privately held companies. Unlike the restricted stock of a public company, the illiquidity of a one-percent block in a private company is similar to the illiquidity of a 30-percent block in the same company. Moreover, the illiquidity of a large block restricted stock transaction is more similar to the illiquidity of any minority interest of a privately held company. Before addressing the statistics in the first quoted paragraph above, the second quoted paragraph should be discussed. Mr. Hall asserts: a) small blocks of public companies are more liquid than larger blocks of public companies; and, b) the illiquidity of a small block in a private company is similar to that of a larger block in the same company. Note first that assertion a) is supported only by the referenced graph. Note second that assertion b) is not supported at all. 117 Hall, Lance S., Block Size vs. The Discount: A Closer Look at the Top Quintile, The Value Examiner, May/June Ibid. Quantifying Marketability Discounts E-Book Page 167

17 We stratified the two year portion of the database by block size. Figure 5-8 provides the results of that stratification, which was sorted by subsequent month pricing discounts. As before, if there was no discount for a transaction in the subsequent month field, we substituted the transaction month discount as a proxy in order to have a complete data set. This could cause some differences between Mr. Hall s calculations and ours. TWO YEAR ANALYSIS Median Restricted Stock Discounts for the Indicated Ranges of Block Sizes Post-Transaction Prior Month Trans. Mo Subsequent Block Size Ranges No. Trans. Pricing Pricing Mo. Pricing Less than 20% 206 Median 16.2% 19.3% 19.0% Average 17.2% 20.2% 20.3% 20% up to 25% 18 Median 19.5% 21.3% 25.4% Average 21.0% 25.8% 32.6% 25% up to 30% 11 Median 23.8% 20.0% 26.6% Average 19.2% 24.8% 28.6% 30% up to 35% 7 Median 36.0% 40.9% 40.3% Average 23.2% 38.1% 41.8% 35% and Over 6 Median 37.7% 48.2% 30.6% Average 35.5% 44.6% 35.8% Total Transactions 248 Median 16.7% 20.0% 22.2% Average 18.2% 21.9% 22.7% Figure 5-8 In the first paragraph quoted in this section, Mr. Hall stated that small percentage blocks (less than 20%) had median discounts of only 18%, whereas the largest blocks (over 35%) had a median discount of 58%. Our analysis essentially affirms the statement regarding small percentage blocks. Our calculations differ somewhat from Mr. Hall s analysis, at least based on subsequent month pricing. Median discounts rise from 25.4% (20% to 25% block sizes) to 26.6% (25% to 30% block sizes), to 40.3% (30% to 35% block sizes), and then fall to 30.6% for block sizes over 35%. What the Hall analysis did not state is that the under 20% segment of the database represents 206 transactions, or more than 80% of all observations. The two segments where Mr. Hall s graph indicates the largest discounts are comprised of a total of only 13 observations, or 5% of the observations. Quantifying Marketability Discounts E-Book Page 168

18 The FMV Discount for Lack of Marketability Method Mr. Hall advances the FMV Approach, which is really a valuation method (within the guideline company method, which is within the market approach to valuation). We summarize the method here in order to illustrate the comparative foundations upon which it is built. Conduct a comparative analysis between the subject interest and the small block portion of the database. Presumably, this would involve comparisons with the 206 transactions having discounts under 20%. Make comparisons on characteristics like revenue, total assets, volatility, book value, market value, profit, profit martins, volatility, market-to-book ratios and dividends. Mr. Hall provides only the first three, plus price per share, in his published analysis. Based on these comparisons, develop an as if restricted stock of a public company discount ( restricted stock equivalent discount or RED) Make another analysis comparing the subject company with the companies exhibiting large discounts. Presumably, such comparisons would be made with the 13 transactions having discounts greater than 30% and develop a large block discount increment, or LBI. Note that the total discount from which the LBI would be calculated (TD) would have to be the result of the analysis of the 13 large block transactions, (although TD is not mentioned specifically in the article). Obtain the appropriate private company discount, (PCD), by summing the RED and the LBI. According to Mr. Hall s article: PCD = RED + LBI We can now see that there is more than a bit of round-about reasoning in the FMV method. It calls for analyzing the small block portion of the database in order to develop RED. It then says to analyze the large block portion to develop LBI, which is the result of a total discount (TD) less the just-derived RED. However, we see that the objective of the analysis, the PCD, is determined once TD is determined. PCD = RED + LBI TD = RED + LBI (per Mr. Hall s definition) (the two add up to the same total) Therefore: TD = PCD Once TD is known, the PCD is also known. And TD is determined by analyzing only a small handful of transactions, as should be clear from Figure 5-8 above. Quantifying Marketability Discounts E-Book Page 169

19 All of this follows from Mr. Hall s unsupported assertion that illiquid interests of private companies, whether large or small, are more like large blocks of restricted shares of public companies (in terms of relative illiquidity) than smaller blocks of restricted shares of public companies. If TD = PCD, and TD is determined from only 13 or so transactions in the data base, why do we need the rest of the database when valuing illiquid interests of private companies? Would it not be better to perform the most in-depth analysis of those 13 transactions possible to obtain the right discount, and then to apply that discount to all situations? In seeking to specify a methodology, the FMV method breaks down. IMPLIED REQUIRED RETURNS, THE QMDM AND THE FMV STUDY The purpose of marketability discount analysis is to determine, to the best of an appraiser s ability, the appropriate discount applicable to each particular valuation situation. The analysis thus far suggests that the FMV Study is of limited use as a basis for guideline company analysis, where the standard is comparability in terms of business characteristics and the proximity of the pricing of guideline transactions with the valuation date. As described in the analysis above, the FMV Study, including both the two year and one year portions, does not provide substantial comparative information that relates to most of private corporate America, at least as observed in our experience. We have suggested that restricted stock studies can be used as evidence of the existence of and general magnitude of restricted stock discounts. We have also suggested that the studies (whether using averages or individual transactions) can be used to infer the implied required rates of return imbedded within the various transactions. 119 Using the quintile analysis of the two year portion of the FMV Study, we can see how this type of analysis works. In the top portion of Figure 5-9 below, the quintile discounts are repeated from Figure 5-5 above. In the middle portion, implied returns for the median transaction in each quintile are shown. In order to estimate a required return, basic present value analysis suggests that three factors are needed: A present value (PV). In this case the discounted price for each median transaction is used (on the basis of remaining percentage of $1.00). A future value (FV). For purposes of this analysis, we have assumed that the expected growth in value (i.e., the discount rate for the public companies assuming no dividends) is 15% per year, compounded, from the base (indexed) price of $1.00 per share. This assumption would not, of course be appropriate for all companies; however, varying the base equity discount rate over a fairly wide range does not change the general conclusions of this analysis. 119 The analysis of this section addresses Mr. Abrams question (noted in Chapter 1) regarding the magnitude of expected holding period premiums implied by the restricted stock studies. Quantifying Marketability Discounts E-Book Page 170

20 An expected holding period (HP). We have estimated the number of quarters needed to dribble out each block under Rule 144 at the greater of 1% of shares outstanding or weekly average trading volume as provided for each transaction in the FMV Study (assuming a constant dribble rate). The implied holding period is then determined by adding the two year period of Rule 144 restriction to the half-life of the years to dribble (quarters / 4 / 2). For clarity, the first quintile holding period of 2.66 years (highlighted below) is calculated as follows: 2 years + (5.25/4)/2 = 2.66 years. Two Year Portion of the FMV Study Transactions First Second Third Fourth Fifth Overall at Premiums Quintile Quintile Quintile Quintile Quintile Medians Restricted Stock Discounts -5.4% 4.9% 15.9% 25.0% 34.7% 51.0% 22.2% Number of Transactions Implied Returns for Expected Holding Periods Based on Rule 144 Dribble-Out Provisions Returns Based on Discounts Assuming a Base Equity Discount Rate of 15.0% for all Companies Prior Month Hi-Lo Avg Pricing 18.4% 18.2% 18.6% 20.5% 26.8% 29.3% 22.4% Transaction Mo. H--Lo Avg 16.2% 19.4% 22.1% 25.2% 32.3% 34.9% 23.9% Subsequent Mo. Hi-Lo Avg 12.3% 17.5% 22.0% 26.9% 34.2% 44.3% 24.4% Quarters to Dribble (Rule 144) Implied Holding Period (Years) Implied Holding Period Premiums (HPP) Relative to Assumed Returns Based on Discounts Base Equity Discount Rate Prior Month Hi-Lo Avg Pricing 3.4% 3.2% 3.6% 5.5% 11.8% 14.3% 7.4% Transaction Mo. H--Lo Avg 1.2% 4.4% 7.1% 10.2% 17.3% 19.9% 8.9% Subsequent Mo. Hi-Lo Avg -2.7% 2.5% 7.0% 11.9% 19.2% 29.3% 9.4% Figure 5-9 The middle portion of the figure displays the median of calculated required returns implied by each quintile s assumptions. For example, the first quintile s subsequent month pricing return of 17.5% is the median of the individual transactions in the quartile calculated as follows: [(1 + 15%)^HP / ( 1 RSD )]^(1/HP) - 1 Note that the implied required returns for the first through the third quintiles are in the range of the high teens to the mid-twenties in percentage points. The implied holding period premiums (relative to the assumed equity discount rate of 15%) are in the range of 3% to 12% or so. The median implied holding period return is 24.4% for the overall two year portion of the study (based on subsequent month pricing), and the median holding period premium is 7.4%. The implied expected returns are lower for the premium transactions and higher for the higher level discounts in the fourth and fifth quintiles. These median expected required returns and expected holding period premiums (R hp and HPP from the QMDM discussion in Chapter 1) appear reasonable in light of the nature of the restricted stock transactions and the general nature of the issuing companies. And the general level of returns can be used as a basis for estimating holding period premiums in the valuation of illiquid interests of private companies. Quantifying Marketability Discounts E-Book Page 171

21 Why might it be a superior methodology to look at restricted stock transactions in terms of their implied required returns rather than in the absolute level of their discounts? First, market participants, particularly actual buyers of illiquid interests in public companies, do not make decisions based on the absolute levels of discounts in other transactions. While the price negotiated will always reflect a discount (positive, zero, or negative), pricing decisions, in our experience, are based on expected return requirements over expected investment horizons. Second, the standard deviation of expected required returns is much smaller than the standard deviation of population restricted stock discounts. The discounts are a single number and reflect only the relationship between current market prices and restricted stock transaction prices. The implied returns, while still a single number, reflect the influence of an estimate of expected return for the public securities, the expected holding period until liquidity, and the transaction price. If the assumptions made are reasonable, the variability of implied required returns should be less than that of raw discounts they take into account more information about the investment decision-making process of investors. The relative variability of restricted stock discounts and implied required returns can be seen in Figure Two Year Portion of the FMV Study Restricted Stock Discounts Implied Required Returns* Prior Mo. Trans Mo. Subs. Mo. Prior Mo. Trans Mo. Subs. Mo. Medians 16.7% 20.0% 22.0% 22.4% 23.9% 25.0% Averages 18.2% 21.9% 22.5% 24.7% 26.5% 27.6% Std Deviations 16.8% 16.0% 21.5% 10.4% 10.6% 13.7% * Assuming base equity discount rate for all companies is 15.0% Figure 5-10 And third, the use of implied required returns in the context of models like the QMDM enables appraisers to simulate the thinking of hypothetical and real life investors when determining marketability discounts. The fact that appraisers must still make estimates of holding period premiums to enterprise equity discount rates when employing the QMDM to value shareholder level interests does not negate the benefit of discounted cash flow analysis. In fact, as indicated in the discussion of the QMDM, appraisers must often estimate specific company and/or specific shareholder-level risks when employing discounted cash flow analysis. And they do so in light of market evidence as outlined in the discussion of the FMV Study. Quantifying Marketability Discounts E-Book Page 172

22 Guideline Analysis Applied to the QMDM Examples We used the QMDM to estimate marketability discounts in three example cases in Chapter 1. The first two examples involved minority interests in partnerships, with one holding non-income producing land, and the other holding an attractive apartment building and providing substantial cash flows to limited partners. The transactions in the FMV Study do not involve any companies holding primarily land, apartments, or securities, so there are no directly comparable observations in the data base. The Business Valuation Resources website contains a list of questions and answers regarding the FMV Study. 120 The question regarding the use of the study in valuing partnerships is addressed there: Q. Is the FMV marketability discount as applicable for a family limited partnership with real estate interest as it is for corporations? R. 1. Since the FMV Restricted Stock Study TM includes data on the differences in value between fully-marketable and less-marketable securities, it could be applied to ANY situation where the marketability discount is an issue, but the data is clearly more applicable in some situations than in others. 2. The data is most applicable to the following marketability discount determinations, in descending order: (a) the restricted (subject to Rule 144) stock of publicly traded companies; (b) the stock of privately held operating entities; (c) everything else. 3. When valuing non-controlling illiquid interest in FLPs holding real estate, analysts often use transactions from the secondary market for partnership interests (Partnership Spectrum). This data, since it pertains directly to real estate holding partnerships, are most directly applicable to RE holding FLPs. However, the lack of marketability discount may not be fully reflected in the secondary market discount data, since these interests do have a market, albeit not a very active or liquid one. Thus, the FMV study could be used to determines a marketability discount for FLPs holding real estate either (a) directly, through determining a discount for entities that are relatively similar to the subject entity and adding this discount to the discount from the secondary market transactions or (b) as a smaller increment for the incremental lack of marketability of an interest in an FLP vs. an interest in a partnership trading in the secondary market. This last discount could be based on the difference between large block and small block transactions in the FMV study (since large blocks are less liquid than small blocks). 120 See Quantifying Marketability Discounts E-Book Page 173